The present invention relates to power conditioning configurations and more specifically to a method and apparatus for modifying a pre-charge voltage on a DC link between a rectifier and an inverter where the rectifier may be fed by more than one source and at least two of the potential sources have disparate characteristic impedances.
Power plants are linked to power consuming facilities (e.g., buildings, factories, etc.) via utility grids designed so as to be extremely efficient in delivering massive amounts of power. To facilitate efficient distribution over long distances power is delivered as low frequency three phase AC current.
Despite being distributable efficiently, low frequency AC current is not suitable for end use in consuming facilities. Thus, prior to end use utility grid power has to be converted to a useable form. To this end a typical power "conditioning" configuration includes an AC-to-DC rectifier that converts the utility AC power to DC across positive and negative DC buses (i.e., across a DC link) and an inverter linked to the DC link that converts the DC power back to three phase AC power having an end useable form (e.g., three phase high frequency AC voltage). A controller controls the inverter in a manner calculated to provide the voltage waveforms required by the consuming facility.
While power plants are typically the most efficient way to provide massive amounts of power to consuming facilities, most consuming facilities are equipped with a backup power supply, typically a generator, to provide power to the facility when utility grid power is cut off. Generators, like power plants, provide three phase AC power to conditioning configurations.
A typical rectifier includes a diode or SCR bridge that, while rectifying AC power, causes a three phase 360 Hertz ripple across the DC bus. To filter out the ripple many conditioning configurations include a pre-charge component, typically a large bulk capacitor, linked between the positive and negative DC buses. Unfortunately, under certain circumstances the bulk capacitor can result in a current in-rush into the inverter that can damage the inverter. The magnitude of the current in-rush is a function of two factors. First, if the bulk capacitor is relatively uncharged the difference between a peak applied voltage and the DC bus voltage can be substantial, the potential difference tending to cause the current in-rush. Second, the in-rush magnitude is also a function of the AC source (e.g., the utility) impedance. Where source impedance is relatively large the in-rush is appreciably tempered and where source impedance is relatively small the in-rush is relatively large.
Because of the potentially damaging current in-rush, most conditioning configurations including a DC bus bulk capacitor also include a hardware protection mechanism to ensure that the bulk capacitor charge is always above a safe threshold or "pre-charge" level prior to causing the inverter to draw power from the DC bus. The protection mechanism may be as simple as a DC bus voltage sensor for sensing the bulk capacitor charge, a comparator to compare the bulk capacitor charge to the pre-charge level and a mechanism for causing the inverter controller to disable (i.e., opening inverter switches so the inverter does not draw power) the inverter when the bulk capacitor charge dips below the pre-charge level. Thus, upon start-up, the DC bus capacitor is pre-charged to the pre-charge level prior to power being drawn by the inverter. Similarly, during inverter operation if, for some reason, the capacitor charge dips below the pre-charge level, the inverter stops drawing power until the capacitor charge again exceeds the pre-charge level.
A typical impedance for a utility source is approximately a relatively low 1 to 3%. Given such a low impedance, a typical pre-charge level is approximately 85% of full bulk capacitor charge or full charge level (FCL). Thus, when the bulk capacitor charge falls below the 85% FCL, the inverter controller disables the inverter. While some conditioning configurations have an adjustable pre-charge level, each configuration requires that the pre-charge level be set to accommodate a specific source impedance.
While a utility source is typically characterized by a low impedance, most generators are characterized by relatively high impedances. For example, many generators have impedances that exceed 20%. Such high impedances often cause excessive distortion in AC supply voltage and current waveforms. Referring to FIG. 2 exemplary voltage and current waveforms V and I, respectively, corresponding to a generator source are illustrated. Waveforms V and I were generated by a generator feeding a six-pulse inverter drive under load. While a generator regulator can increase output to provide a desired RMS, the distorted waveforms can cause DC bus voltage dips.
In many cases, when the DC bus voltage dips below the pre-charge level, the dip causes pre-charge disablement of the inverter while the bus voltage is re-charged to the pre-charge level. Because pre-charge levels are set for efficient and safe operation assuming a utility source (e.g., 85% of full bulk capacitor charge), and generators often cause relatively large DC bus voltage distortion, switching from a utility source to a generator source often results in the DC bus voltage dropping below the specified pre-charge level and hence causes the controller to disable the inverter during pre-charge sessions.
One solution to the voltage dip problem caused by switching from a utility to a generator source is to provide a much larger generator (e.g., 3 to 4 times larger) to lower the source impedance to a level more similar to the utility source impedance. While adopted by several members of the industry this solution is relatively expensive and therefore is not suitable for all applications.
Another solution that can be adopted when the pre-charge level is manually adjustable is to adjust the pre-charge level manually to accommodate the generator's higher characteristic impedance. To optimally manually adjust the pre-charge level as a function of source and system characteristics a system user has to be knowledgeable about many system characteristics and inductive/electrical phenomenon. For example, to optimally tune the user has to be familiar with an expected peak source voltage, source impedance, inverter switch characteristics such as current handling capabilities and so on. In other words, the user has to be relatively highly skilled to effect optimum pre-charge tuning. While effective, manually adjustable systems are burdensome in most cases for several reasons. First, generators are primarily used as backup power sources and therefore are only sporadically employed. For this reason often no user at a consuming facility is familiar with the pre-charge level adjustment procedure and the source/system characteristics described above. Second, even when generators are employed the periods of employment are often limited as utility sources are typically only cut off for relatively short times. Thus, even if the pre-charge level could be manually adjusted at a consuming facility when a generator is employed, the manual adjustment task would likely be for naught when the utility source is back on line. In fact, in such a case, when the utility is back on line the pre-charge level would again have to be adjusted back to the level corresponding to the utility impedance.
Other solutions such as halting power consumption when a utility is down or simply allowing dips and pre-charge sessions when a generator source is employed are unacceptable in many applications where relatively constant power and inverter operation is required.
In addition to consuming facilities that are primarily powered by utility sources, some remote facilities are powered primarily by dedicated generator sources. In these cases the conditioning configurations that facilitate manual pre-charge level adjustment are advantageous as the pre-charge level can be set to accommodate specific generator characteristic impedance. Unfortunately, even in this case, the task of selecting the pre-charge level is burdensome and increases the time and costs associated with commissioning the conditioning configuration.
Therefore, a need exists for a low cost system that facilitates switching of power sources that feed conditioning configurations where the sources are characterized by different impedances without causing unnecessary inverter disablement. In addition, it would be advantageous to have a system that automatically selects a suitable pre-charge limit as a function of the source linked to the conditioning configuration.